Focal adhesion kinase regulates insulin resistance in skeletal muscle
- First Online:
- Cite this article as:
- Bisht, B., Goel, H.L. & Dey, C.S. Diabetologia (2007) 50: 1058. doi:10.1007/s00125-007-0591-6
On the basis of our previous studies, we investigated the possible role of focal adhesion kinase (FAK) in the development of insulin resistance in skeletal muscle, a major organ responsible for insulin-stimulated glucose uptake.
Materials and methods
Insulin-resistant C2C12 skeletal muscle cells were transfected with FAK wild-type or FAK mutant plasmids, knocked down using small interfering RNA (siRNA), and their effects on the levels and activities of insulin-signalling molecules and on glucose uptake were determined.
A significant decrease in tyrosine phosphorylation of FAK in insulin-resistant C2C12 cells was observed. A similar decrease was observed in skeletal muscle obtained from insulin-resistant Sprague–Dawley rats fed a high-fat diet. Increased levels of FAK in insulin-resistant C2C12 skeletal muscle cells increased insulin sensitivity and glucose uptake. These effects were reversed by an increase in the level of kinase activity mutant FAK or suppression of endogenous FAK by siRNA. FAK was also found to interact downstream with insulin receptor substrate-1, phosphatidylinositol 3-kinase and protein kinase C and glycogen synthase kinase 3β, leading to translocation of glucose transporter 4 and resulting in the regulation of glucose uptake.
The present study provides strong evidence that the modulation of FAK level regulates the insulin sensitivity of skeletal muscle cells. The results demonstrate a direct role of FAK in insulin-resistant skeletal muscle cells for the first time.
KeywordsFAK Insulin resistance siRNA Skeletal muscle
focal adhesion kinase
glycogen synthase kinase 3β
MCDB 201:Ham’s F-12 serum-free medium (1:1)
MCDB 201:Ham’s F-12 serum-free medium with insulin (1:1)
phosphatidyl inositol 3-kinase
protein kinase C
small interfering RNA
Focal adhesion kinase (FAK) is a non-receptor tyrosine kinase and is known to play a key role in integrin signalling . FAK can associate with multiple cellular components, including adhesion-associated proteins and signalling molecules . We have previously reported regulation of myogenesis [3, 4] and cell spreading [5, 6] by FAK as a function of its tyrosine phosphorylation in C2C12 skeletal muscle cells due to insulin stimulation. Bianchi et al.  reported the regulation of FAK phosphorylation by glycogen synthase kinase 3β (GSK-3β) during cell spreading and migration. In addition, regulation of FAK phosphorylation by insulin stimulation in non-muscle cells has been reported [8–10]. A regulatory role of FAK in IRS-1 levels [11, 12] and hepatic insulin action [13, 14] has also been reported. Genetic studies have provided evidence for interaction between integrin and insulin signalling [15, 16]. The available literature also focuses on the role of FAK in insulin-mediated effects on the cytoskeleton [10, 17]. Recently, Huang et al.  have reported that a reduced level of FAK disrupts insulin action in skeletal muscle. However, the role of FAK in regulating insulin resistance in skeletal muscle, if any, and its possible pathway have not been investigated.
We had previously generated an insulin-resistant skeletal muscle model by differentiating mouse C2C12 cells in the chronic presence of insulin (100 nmol/l) in serum-free medium (MFI) [19–22]. The model was validated using glucose-lowering drugs, including pioglitazone , metformin  and gliclazide . This model was used in the present study to investigate the role of FAK, if any, under conditions of impaired insulin action.
Materials and methods
Nutrient Mixture F-12 Ham, MCDB 201 medium and 2-deoxy[3H]glucose (2-DOG) were from Sigma Chemical Company (St Louis, MO, USA). Anti-phospho Akt (Ser473), anti-Akt, anti-phospho (Ser) GSK-3β and anti-GSK-3β antibodies were from Cell Signaling Technology (MA, USA). Anti-FAK, anti-phosphotyrosine, anti-insulin receptor (IR) and anti-IRS-1 antibodies were from Santa Cruz Biotechnology (Santa Cruz, CA, USA). Bovine insulin and anti-GLUT4 (also known as SLC2A2) was purchased from Calbiochem (CA, USA). All other reagents were from Sigma unless attributed otherwise.
C2C12 skeletal muscle cell lines (wild-type and transfectants) were maintained in DMEM supplemented with 15% FCS containing penicillin 100 IU/ml and streptomycin 100 μg/ml in 5% CO2 at 37°C . Cells at 70% confluence were differentiated in an equal mixture of MCDB 201 and Ham’s F-12 medium as described previously  in the absence (MF, insulin-sensitive condition) or in the chronic presence of 100 nmol/l insulin (MFI, insulin-resistant condition) for 3 days.
C2C12 myotubes, pretreated with or without insulin, were washed with ice-cold PBS, scraped into buffer A (mmol/l: HEPES 20, pH 7.4, sucrose 250, EDTA 1, phenylmethylsulphonyl fluoride [PMSF] 2) with 10 μg/ml aprotinin and leupeptin and homogenised for 10 min at 4°C, after which low-density microsome (LDM) and plasma membrane fractions were isolated by differential centrifugation . Protein samples (100 μg) were subjected to SDS-PAGE and western immunoblotting.
Transfection of plasmid vectors
C2C12 cells were stably transfected with FAK wild-type and mutant plasmids as described earlier .
siRNA oligonucleotides against FAK targeted sequence 5′-TGCAATGGAACGAGTATTAAA-3′ were designed and synthesised by Qiagen GmbH (Hilden, Germany). C2C12 cells were transiently transfected with FAK-specific and non-specific siRNA (400 nmol/l) using RNAifect transfection reagent (Qiagen) in reduced-serum Opti-MEM media (Gibco BRL, Carlsbad, CA, USA) according to the manufacturer’s instructions, with minor modifications.
Preparation of cell lysates
Cells were washed with ice-cold PBS and lysed in lysis buffer (mmol/l: HEPES 50, pH 7.4, NaCl 150, MgCl2 1.5, EGTA 1, sodium pyrophosphate 10, sodium fluoride 50, β-glycerophosphate 50, Na3VO4 1, PMSF 2) with 1% Triton X-100, 10 μg/ml each of leupeptin, aprotonin and soyabean trypsin inhibitor at 4°C for 30 min. Homogenates were centrifuged at 16,000 g for 15 min at 4°C . Protein estimation was performed using the Bicinchoninic Acid Kit (Sigma) according to the manufacturer’s instructions. Equal concentrations of lysates of all the samples were subjected to SDS-PAGE followed by western immunoblotting .
Tissue preparation from rats fed a high-fat diet
Male Sprague–Dawley rats (Central Animal Facility, NIPER, India) were divided into two equal groups. One of the groups was fed a normal pellet diet, while the other group was fed a high-fat diet for 4 weeks. The high-fat diet comprised: powdered pellets (364 g), lard (310 g), casein (250 g), cholesterol (10 g), dl-methionine (3 g), Yee-sac powder (1 g), vitamin and mineral mix powder (60 g), sodium chloride (2 g). Biochemical analyses and tolerance tests were performed as described previously [24, 25]. Skeletal muscle from the thigh region was collected, washed with PBS (pH 7.4) and lysed using the lysis buffer as described earlier . The use of animals for the experiments was in compliance with the guidelines of our institutional animal ethics committee.
Phosphatidyl inositol 3-kinase activity
Cells were differentiated under MF and MFI conditions and were stimulated with insulin for 10 min. Cells (500 μg) were lysed and immunoprecipitated for IRS-1 . IRS-1-associated activity of phosphatidyl inositol 3-kinase (PI3K) was determined with the PI3K ELISA Kit (Echelon Biosciences Inc., Salt Lake City, UT, USA) according to the manufacturer’s instructions and as reported previously [26–28].
Glucose uptake assays
Glucose uptake assays were performed using 2-DOG as described earlier .
Relative densitometric analyses of the samples were determined using Gel Doc 2000 (Bio-Rad, Hercules, CA, USA) and Quantity One 1-D analysis software (Bio-Rad) by giving an arbitrary value of 1.0 to the respective control samples of each experiment, keeping the background value as 0.
The data are expressed as mean±SE. For comparison of two groups, p values were calculated using the two-tailed unpaired Student’s t test. In all the cases p < 0.05 was considered to be statistically significant.
Transfected cells were then subjected to the MF and MFI conditions and tyrosine phosphorylation of FAK was assayed (Fig. 2b). Under both conditions, tyrosine phosphorylation of C2FAKwt/+ cells increased by 56 ± 0.06 and 42.2 ± 0.03% over their respective controls (Fig. 2b, lane 3 vs lane 1, p < 0.01; lane 4 vs lane 2, p < 0.01). Although tyrosine phosphorylation of FAK in the MFI condition was always less than in the MF condition, insulin-resistant C2FAKwt/+ cells exhibited increased tyrosine phosphorylation, and the level of phosphorylation was similar to that of insulin-sensitive C2FAKwt cells (Fig. 2b, lane 4 vs lane 1). Tyrosine phosphorylation of FAK in C2FAKmut/+ cells under the MF or MFI condition was less (by 29.4 ± 0.06 and 25.8 ± 0.05%, respectively) than in their respective controls (Fig. 2b, lane 5 vs lane 1, p < 0.01; lane 6 vs lane 2, p < 0.05). The data therefore suggest correlation between FAK phosphorylation and insulin resistance.
To assess the functional contribution of altered FAK protein level and/or its activation in the pathogenesis of insulin resistance, glucose uptake was measured in the above-mentioned cell lines cultured under the MF or MFI condition. Insulin-sensitive C2FAKwt/+ cells showed a 40.5 ± 0.48% increase in insulin-stimulated 2-DOG uptake compared with C2FAKwt cells (Fig. 2c, lane 3 vs lane 1, p < 0.01). For detailed data with and without insulin stimulation see ESM Fig. 3. Under insulin-resistant conditions, C2FAKwt/+ cells exhibited increased glucose uptake (28.6 ± 0.61%) compared with C2FAKwt cells (Fig. 2c, lane 4 vs lane 2, p < 0.01), comparable with that observed in insulin-sensitive C2FAKwt cells (Fig. 2c, lane 4 vs lane 1). Cells with increased levels of FAK were thus protected from the development of hyperinsulinaemia-induced insulin resistance. However, C2FAKmut/+ cells developed insulin resistance as there was no insulin-stimulated increase in glucose uptake (Fig. 2c, lane 5 vs lane 1, p < 0.01 and lane 6 vs lane 2, p < 0.01).
We next examined the potential molecular mechanisms that might underlie the protection from insulin resistance conferred by an increased level of FAK and corresponding tyrosine phosphorylation. The possible mechanisms include increased insulin-mediated activation of the IR and/or its downstream molecules. Therefore, we examined IR-β tyrosine phosphorylation and production in C2FAKwt, C2FAKwt/+, C2FAKmut/+ and C2siRNAFAKwt cells under the MF and MFI conditions. No significant change in insulin-mediated activation or level of IR-β due to modulation of FAK activation was observed (data not shown).
To explore the role(s) of FAK in regulating downstream molecules, IRS-1-associated PI3K activity was determined. In C2FAKwt/+ cells, insulin-stimulated PI3K activity increased 1.7 ± 0.29-fold in the MF condition (Fig. 4c, lane 4 vs lane 2, p < 0.01), whereas it decreased 2.18 ± 0.08-fold in C2FAKmut/+ cells (Fig. 4c, lane 6 vs lane 2, p < 0.01), and in the mutant cells the PI3K activity was same as that of insulin-resistant C2FAKwt cells (Fig. 4c, lane 6 vs lane 8). C2FAKwt/+ cells under the MFI condition showed a 3.6 ± 0.25-fold increase in PI3K activity (Fig. 4c lane 10 vs lane 8, p < 0.01), and the extent of activation was same as that of insulin-sensitive C2FAKwt cells (Fig. 4c, lane 10 vs lane 2). Insulin-stimulated IRS-1-associated PI3K activity was unaltered in C2FAKmut/+ cells under the MFI condition (Fig. 4c, lane 12 vs lane 11), but decreased 4.4 ± 0.01-fold and 1.2 ± 0.01-fold under the MF and MFI conditions, respectively (Fig. 4d, lane 6 vs lane 2, p < 0.01; lane 8 vs lane 4, p < 0.01) in C2siRNAFAKwt cells.
PI3K activates several PIP3-dependent serine/threonine kinases, e.g. Akt (also known as protein kinase B [PKB]) and protein kinase C (PKC) [34, 35]. Therefore, insulin-stimulated activation, if any, of Akt was examined in all the cell lines (Fig. 4e). No additional insulin-stimulated effect of Akt phosphorylation (Ser473) was observed in C2FAKwt/+ cells (Fig. 4e, lane 4 vs lane 2). This may have been due to saturable activation of Akt. Although under the insulin-resistant condition C2FAKwt/+ cells showed additional insulin-stimulated Akt phosphorylation, the extent of activation was not equivalent to that seen in insulin-sensitive C2FAKwt cells (Fig. 4e, lane 10 vs lane 2). C2FAKmut/+ cells displayed decreased Akt activation in both the MF and the MFI condition (Fig. 4e, lane 6 vs lane 2, p < 0.01; lane 12 vs lane 8, p < 0.01). Similar decreases were observed in C2siRNAFAKwt cells (66.2 ± 0.2 and 58 ± 0.13% in the MF and MFI conditions, respectively) (Fig. 4f, lane 6 vs lane 2, p < 0.01; lane 8 vs lane 4, p < 0.01). This indicates possible signal divergence under the insulin-resistant condition.
PKC is reported to regulate glucose uptake downstream of PI3K [34, 35]. Therefore, insulin-stimulated activation of PKC, if any, was tested in all the cell lines by probing with anti-phospho pan-PKC antibody. Insulin-stimulated PKC phosphorylation of C2FAKwt/+ cells increased (by 21.4 ± 0.05%) under the MF condition (Fig. 4g, lane 4 vs lane 2, p < 0.01) but decreased (by 37.4 ± 0.03%) in C2FAKmut/+ cells (Fig. 4g, lane 6 vs lane 2, p < 0.01), and the level of activation was similar to that in insulin-resistant C2FAKwt cells (Fig. 4g, lane 6 vs lane 8). Insulin-stimulated PKC phosphorylation increased (76.3 ± 0.03%) in C2FAKwt/+ cells under the MFI condition (Fig. 4g, lane 10 vs lane 8, p < 0.01), and the extent of activation was as observed in insulin-sensitive C2FAKwt cells (Fig. 4g, lane 10 vs lane 2), whereas it was further decreased (37.5 ± 0.01%) in C2FAKmut/+ cells (Fig. 4g, lane 12 vs lane 8, p < 0.01) without affecting its protein level (Fig. 4g). PKC phosphorylation of C2siRNAFAKwt cells decreased by 62.6 ± 0.02 and 56 ± 0.01% under the MF and MFI conditions, respectively (Fig. 4h, lane 6 vs lane 2, p < 0.01 and lane 8 vs lane 4, p < 0.01) without altering the protein level. The data confirm the involvement of PKC and suggest its possible involvement in signal divergence in FAK-mediated insulin signalling.
In the present study a direct correlation was observed between FAK protein level/phosphorylation and glucose uptake in the presence of insulin, suggesting insulin-dependent regulation of insulin resistance in C2C12 skeletal muscle cells by FAK. Huang et al.  reported that downregulation of FAK affects only insulin-stimulated glucose uptake, unlike the observations in our study (Fig. 3b). One of the possible explanations for this could be that Huang et al. reported glucose uptake when there was 53% downregulation in the level of FAK, leaving approximately 47% of the endogenous FAK available, whereas in our study 85% of the FAK was downregulated, thus eliminating any significant contribution of the endogenous protein. Under our experimental conditions, the mechanism of FAK-mediated regulation of insulin resistance could be the interaction of FAK with unidentified substrate(s) that becomes activated only in the presence of insulin and is necessary for the interaction between FAK and insulin signalling molecules. The observed increase in tyrosine phosphorylation in the basal condition of overexpressed FAK might be due to activation by other stimuli.
Like us, Annabi et al.  and Huang et al.  also observed no interaction between FAK and the IR in their studies, using CHO and HepG2 cells. However, we observed that up- or downregulation of FAK activity correspondingly affected insulin-stimulated glucose uptake via the IRS-1/PI3K pathway. Therefore, our results indicate that FAK regulates insulin signalling upstream of IRS-1 and then regulates the cascade downstream via PI3K. Insulin signalling without involving the IR, via IRS-1, has been reported previously in other non-receptor tyrosine kinases, pp59Lyn in 3T3-adipocytes . FAK is also reported to interact with cytosolic tyrosine kinase Src, which can regulate IRS-1 tyrosine phosphorylation in CHO cells . In human embryo kidney cells, Lebrun et al.  have reported that FAK can phosphorylate IRS-1 without the participation of Src.
In our study, the mechanism whereby mutant FAK inhibits the ability of insulin to induce phosphorylation of Akt remains unclear. It is possible that endogenous FAK interacts with unidentified protein(s) necessary for interaction with Akt. Mutant FAK possibly exerts a dominant negative effect on endogenous FAK. A similar dissociation between these two proteins has been reported previously in hepatocytes . However, an increased level of FAKwt did not restore Akt phosphorylation under insulin-resistant conditions even though tyrosine phosphorylation of IRS-1 and PI3K activation was restored. A possible explanation includes additional regulation of Akt independently of PI3K in the insulin-resistant condition. Lin et al.  found that PKC activator decreased the ability of insulin to phosphorylate both Akt and GSK-3β in rat skeletal muscle, whereas PI3K activation was unaffected.
Decreased glucose uptake irrespective of an increased or decreased level of FAK under conditions of PKC inhibition indicates a markedly smaller contribution of Akt. Therefore, in the insulin-resistant condition the signal travels mainly through activated PI3K to PKC rather than to Akt. Earlier reports have suggested that phosphatidylinositol (3,4,5)-triphosphate (PIP3) produced by PI3K, responsible for both Akt and PKC phosphorylation, has more affinity for PKC than Akt . Moreover, in the insulin-resistant condition PKC downregulates Akt and this is unaccompanied by a change in PI3K activity in skeletal muscle cells . Therefore, it is possible that FAK differentially regulates insulin signalling under normal and insulin-resistant conditions. Sajan et al.  recently reported a differential role of the atypical PKC isoforms ζ and λ in insulin-stimulated glucose transport. Regulation of GLUT4 translocation in skeletal muscle of high-fat-diet-fed rats by PKC (ζ/λ) has also been reported . Therefore, there is a possibility of FAK-regulated activity of PKC ζ/λ, or any other isoform(s), in regulating insulin resistance. Further studies are in progress to determine the specific isoform(s) that may be responsible for interacting with FAK to regulate the pathway.
It has been reported that modulation of FAK level in hepatocytes regulates glycogen synthesis [7, 14]. We also observed that up- or downregulation of FAK correspondingly phosphorylates GSK-3β, implying the regulation of glycogen synthesis by FAK in these cells.
Kotliar and Pilch  observed that the increased level of GLUT4 fails to reconstitute insulin-stimulated glucose transporter in C2C12 cells and postulated the requirement for additional gene(s) to form functional translocating vesicles. Microtubules have been reported to be required for the translocation of GLUT4 [18, 42–46]. FAK is required for the stabilisation of microtubules in fibroblasts  and its tyrosine phosphorylation is essential for the insulin-stimulated rearrangement of actin fibres in CHO/IR cells . Huang et al.  reported that FAK regulates insulin-mediated cytoskeletal rearrangement essential for normal glucose transport and glycogen synthesis. Therefore, the observed increase in glucose uptake as a result of increased FAK level with no change in expression in our study is possibly due to the enhanced recruitment of actin filament and microtubules responsible for the GLUT4 translocation. Further studies are in progress to demonstrate interaction between GLUT, microtubules and FAK.
In summary, the present study demonstrates a novel role for FAK as a positive regulator of the insulin signalling pathway, leading to glucose transport and insulin sensitisation. This conclusion is based on the following findings: (1) in the insulin-resistant condition FAK was always found to be less active; (2) downregulation of FAK activity resulted in impaired insulin signalling; (3) decreased glucose uptake was a consequence of impaired insulin signalling due to downregulation of FAK; and (4) increased level of FAK in the insulin-resistant condition reversed all the above-mentioned observations. The stimulatory effect of FAK on insulin-stimulated glucose transport involves the IRS-1/PI3K/PKC arm of the insulin signalling pathway. Interestingly, IR and Akt do not seem to contribute to the pathway. Since an increase in FAK and its phosphorylation sensitises C2C12 skeletal muscle cells to insulin stimulation of GLUT4 translocation and glucose transport, these results further support a role for an insulin-stimulated FAK/IRS-1/PI3K/PKC signalling pathway as a physiologically important mediator of the action of insulin. Since increased levels of FAK leads to cellular insulin sensitisation, these results also raise the possibility that interfering with FAK production or activation may provide opportunities for developing new therapeutic strategies against insulin resistance. Chemical activators of FAK would be expected to act as insulin sensitisers, which could have beneficial effects in insulin resistance induced by hyperinsulinaemia. This opens up the possibility of new therapeutic intervention.
We thank C. L. Kaul, formerly director of NIPER, for his keen interest in this study. We also thank P. Ramarao, director of NIPER and head of the Pharmacology Department, for providing us with rat muscle samples. We gratefully acknowledge D. Schlaepfer of the Scripps Research Institute, La Jolla, CA, USA for providing pcDNA3.1-HA-FAKwt and pcDNA3.1-HA-FAKmut plasmids as a kind gift. R. Singh is acknowledged for his assistance in the laboratory. B. Bisht is the recipient of a Senior Research Fellowship and H. L. Goel was the recipient of a Senior Research Fellowship from the Council of Scientific and Industrial Research, Government of India, New Delhi. This study was also supported by a grant from the Department of Biotechnology, Government of India, New Delhi to C. S. Dey (BT/HRD/34/04/2004, dated 30.4.2004; BT/PR3994/MED/14/498/2003, dated 30 September 2004).